Background Pedestrian injuries are a leading cause of paediatric injury. One aspect of pedestrian safety is selection of safe routes across roads. Children often select routes based on expedience but not safety.

Aims Test efficacy of video-based training on safe pedestrian route selection for children ages 7–8.

Methods As part of a larger study, 240 children were randomly assigned to one of four groups: video training, virtual reality (VR) training, streetside training, control. Videos included instructions on selecting safe crossing routes; other groups received training in different aspects of pedestrian safety (VR, streetside) or no training (control). All training groups received six 30-minute sessions.

A story task evaluated route selection skills before and after interventions. Children heard four brief vignettes accompanied by graphic representations of streetside locations, then selected preferred crossing routes from three options: distant crosswalk (safest), directly across street without crosswalk (less safe), diagonally across intersection (riskiest).

Results Among children trained by video, the percentage selecting the safest route in all scenarios increased 56% post-training. For VR, streetside training and control groups, comparable increases were 9%, 19%, −33%. One-way ANCOVA (IV: condition; DV: post-training score; covariates: age, pre-training scores) yielded significant results, F(3211)=3.46, p<0.05. Post-hocs confirmed video group (M=1.85) scored safer post-training than VR (M=2.13, p<0.01) or controls (M=2.06, p<0.05), plus trend with streetside training (M=2.01, p=0.09).

Significance Video education apparently teaches children safer pedestrian route selection than control groups. It is inexpensive and easy-to-implement, and warrants broad dissemination.